JPH06243210A - Image processing method and device therefor - Google Patents

Abstract

PURPOSE:To ensure the effective use of an image processor containing a scanner and a printer by connecting the image processor to a communication network. CONSTITUTION:A server 102 is connected to a network via an Ethernet port 219 and also to a copying machine 103 via the video interface ports 221 and 222. The server 102 can recognize, encode and translate the characters for the image data read by the machine 103 or acquired from the network through a character recognizing circuit 210, an image coding/decoding circuit 211 and an automatic translating circuit 213. Furthermore the image data read by the machine 103 are converted into a PDL form by a PDL converting circuit 212 for reduction of the data quantity. Then the reduced image data can be sent to a host device.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus and method for controlling input / output of a scanner / printer in which an image scanner and a printer are integrated.

[0002]

2. Description of the Related Art In recent years, networking of system resources has advanced, and LAs have been installed in intelligent buildings as a whole.
A large-scale network in which N is distributed is realized. Furthermore, a nationwide network in which a LAN is directly connected to a public line such as WAN (Wide Area Network) and an advanced information network such as ISDN are being developed.

Therefore, the user can use a computer on a floor or a building different from the place where the user is located, and further, it can be separated from a user in Tokyo at a long distance in Osaka or the like. It is becoming possible to use host computers that have Along with that, the printer server also
In addition to the conventional use in a relatively narrow range, it has become possible to use in a very wide area.

Conventionally, in order to share a printer as a network resource on the network, it has been realized to connect the printer to the network via a printer server. However, I never shared a scanner to read images.

[0005]

When inputting using a scanner, all input data are treated as bitmap data. Therefore, when the scanner is used as a shared resource on the network, the scanned data goes through the network as image data, which not only reduces the network efficiency but also manages the network files. A lot of memory area was required even when handling it with the file server. Also,
Since it is bitmap data, even if the read original is a document, it is not possible to perform operations such as editing as a document.

Therefore, even if a scanner-printer including a scanner and a printer is connected to a network, the image data read by the scanner must be appropriately processed.
There is a problem that the scanner cannot be effectively used as a shared resource of the network.

The present invention has been made in view of the above-mentioned conventional example, and various image processing services can be used, and the image processing apparatus and method can be efficiently used by reducing the amount of image data read from the scanner. The purpose is to provide.

[0008]

[Means for Solving the Problems] and

In order to achieve the above object, the image processing apparatus of the present invention has the following configuration.

Input means for inputting image data, sorting means for sorting the image data input by the input means according to the type of image contained therein, and an image according to the type of image recognized by the sorting means Coding means for coding the data and conversion means for converting the image data and the code data obtained by the coding means into a page description language are provided.

The image processing method of the present invention has the following configuration.

An input step of inputting image data, a segmenting step of segmenting the image data input by the input step according to the type of image contained therein, and an image according to the type of image recognized by the segmenting step. A coding process for coding the data and a conversion process for converting the image data and the code data obtained by the coding process into a page description language are provided.

[0012]

【Example】

[Embodiment 1] As a preferred embodiment of the present invention, a network in which a copying machine is connected as a scanner printer will be described.

FIGS. 1 and 2 show a part of the network of this embodiment. <Configuration of Network and Copier (FIG. 1)> FIG. 1 is a schematic diagram of a network system of the present embodiment. A host computer 101 and a copying machine 103 are connected to a network 104 via a scanner / printer server 102. The host computer 101 executes a client process 105 for controlling the copying machine 103. Further, in the scanner printer server 102,
Under the control of the client process 105, the copying machine 1
The server process 106 that controls 03 is running. The client process 105 uses the Ethernet 10
The image is read from the scanner side of the copying machine 103, and the printer side prints the image by communicating with the server process 106 via the server 4. In addition, the copying machine 10
Copying within 3 is also possible. Needless to say, the copying machine 103 has therein a scanner for reading an image and a printer for printing an image, and as described later,
The scanner and printer can be used as independent input / output devices. Further, it is possible to use only the functions provided inside without using the scanner and printer of the copying machine.

Scanner printer server 102 and copying machine 1
03, SP as a control signal and a data signal.
COM (scanner printer serial command signal) 10
7, CLOCK (clock signal) 108, VSYNC
(Vertical sync signal) 109, HSYNC (Horizontal sync signal)
110 and DATA (image data signal) 111. These signals are collectively called a video I / F signal (abbreviated as video I / F). The copying machine can be used when the copying machine 103 is used as a scanner for reading image data, when the image data sent as a printer is used for print output, or when an image read by the scanner as a copying machine is used as a printer. There is a case where it is output by, or a case where it is used as a server without input / output. The proper use is based on the instruction given by the SPCOM signal 107, and the video I / F signal is used for inputting / outputting image data and a synchronizing signal.

<Configuration of Scanner Printer Server (FIG. 2)
> FIG. 2 is a block diagram of the scanner printer server 102. Each block in the figure will be described with reference to the figure. In addition,
The scanner printer server may be abbreviated as a server.

[Communication between server / client] When the scanner / printer server 102 is started, its CPU 20
1 starts the program stored in the PROM 202,
The server process 10 using the RAM 203 as a temporary storage location
6 is executed. At this time, the scanner printer server 10
2 can be connected to the Ethernet 104 via the Ethernet port 219 by the Ethernet controller 206 and can communicate with the client process 105 of the host computer 101.

[Communication between server / copier] The scanner / printer server 102 and the copier 103 are CLOCK10.
8, data is exchanged by a video I / F signal such as DATA111. Data exchange by video I / F signals is a completely synchronous type, and SP is used to start or pause or change parameters on the copying machine 103 side.
This is performed by a COM (scanner printer command) signal 107.

The sub CPU 214 is a dedicated CPU that constantly monitors the SPCOM signal 107 between the copying machine 103 and the server 102. This CPU 214 is the copying machine 10.
3 is in continuous communication with 3 to receive block management of image data and errors.

The serial interface 215 serially communicates commands between the scanner / printer server 102 and the copying machine 103 via the video I / F command port 221.

Video I / F signal control circuit 216
Is a circuit for controlling the input / output of the synchronization signal shown in FIGS. 11 and 12, the input of the image data output from the scanner side in synchronization therewith, and the output of the image data sent to the printer side. . The input / output signal controlled there is the video I / F signal described above.
Input / output is performed via the F data port 222.

The dual port RAM 208 includes a CPU 201 for managing the server process 106 and the copying machine 103.
The communication with the sub CPU 214 that manages the interface with the sub CPU 214 is performed by sharing the memory and controlling the interrupt. The sub CPU 214 constantly communicates with the copying machine 103. However, the main CPU 201 only operates when necessary.
Interrupt 214, dual port RAM 208
Pass data through. On the contrary, the sub CPU
When the 214 wants to inform the main CPU 201 of scan end, print end, or error information, the main CPU 2 is sent through the dual port RAM 208.
Interrupt 01 to exchange data.

Scanner printer server 102 and copying machine 1
As a command exchanged with the printer 03 from the scanner printer server 102 to the scanner side of the copying machine 103,
A prescan command, a scan command, etc. are sent. Further, the scanner side of the copying machine 103 sends status information such as a copy command and operation abnormality.

Similarly, a print command or the like is sent from the scanner printer server 102 to the printer side of the copying machine 103. Further, the printer side of the copying machine 103 sends status information such as paper out, paper jam, and operation abnormality.

[Buffer Memory] The image data is temporarily stored in the buffer memory 209 which can store several lines. The buffer memory 209 is accessible from the CPU 201, and character recognition, encoding, and decoding of image data are performed by reading the data from the buffer memory 209.

[Character Recognition Circuit] The character recognition circuit 210 is
A process of converting character image data into character code data is performed.

The conversion procedure performed here is as follows.

(1) Character Cutout Character cutout is a process in which the CPU 201 detects pixel breaks in the vertical and horizontal directions of the data in the buffer memory 209 to detect character blocks. Actually, a certain threshold value process is applied to the touch image data (halftone image) read by the scanner 103 to determine the presence or absence of pixels. For example, if the character image data is "Today is blue sky", the character is cut out and the "book",
"Sun", "ha", ",", "clear", "heaven", "na",
It is cut out into a character section called "ri".

(2) Transfer for each character section The CPU 201 reads out image data of a character section from the buffer memory 209 and sends it to the character recognition circuit 210.

(3) Feature Extraction of Character Section The extracted character section is divided into finer blocks, and in each of the divided blocks, the pixel sequence is left, right, up, and down.
Detects which direction it is facing diagonally.

(4) Similarity calculation Based on the feature extraction information in the fine blocks of the feature extraction (3), characters having a high feature similarity are extracted from the dictionary. A detailed block diagram is shown in FIG. 6 described later.

The read image is recognized as a character by the above four-step procedure.

[Image Encoding / Decoding Circuit] The encoding / decoding circuit 211 encodes the data read from the scanner of the copying machine 103. Also, the encoded image data sent from the Ethernet 105 is decoded.
As this encoding method, ADCT (Adactive Discret
eCosine Trans-form) method. The ADCT coding method is a color still image coding method and is described by the block diagram of FIG.

[PDL conversion circuit] PDL conversion circuit 212
Is a means for converting an image developed in the buffer memory 209 with a bit map or a byte map into a PDL (Page Description Language) for describing a print command to the printer. PDL conversion circuit 2
The block diagram of 12 is according to FIG. 9 described later.

[Automatic Translation Circuit] The characters coded by the character recognition circuit 210 are put together for each clause to form one sentence. The automatic translation circuit 213 receives the one sentence as data and translates it into a language designated by the host computer. For example, the language read by the scanner and recognized is Japanese, and it is translated into English.

The following steps are taken as a translation procedure.

(1) Morphological analysis Words in one sentence are separated.

For example, in the case of "I go to my friend's house today,""I","wa","today","friend","no","home","ni" , "Play", "To",
It is divided into "go" and ".".

(2) Interpolation analysis Modifiers are detected from the words subjected to morpheme analysis.

For example, a word that specifically modifies “house” is detected as “friend” → “no” → “house”.

(3) Semantic analysis The consistency is checked from the context of the sentence.

(4) Replacement Rule Analysis In the case of Japanese to English, it is converted into an English rule. In the case of Japanese, the subject is first or first and the predicate is last, but in English, there are certain rules. Perform that conversion.

For example, rearranging "I go to a friend's house to play" in the order of English words, "I am""go"
It becomes like "to play,""to,""house of friends." This is also the case between English and Japanese, or other two languages, if the rules are replaced.

(5) Translation Perform translation such as Japanese-English conversion and English-Japanese conversion based on the replacement rule.

A language can be translated into another language by the above-mentioned five-step procedure.

The automatic translation circuit 213 may be provided in the server 102 or in the host computer. Here, the automatic translation circuit 213 is provided in the server 102.

[Hard Disk] The HD (hard disk) 226 is used for temporarily storing print data (code and image data) transferred from the network during printer processing. By temporarily storing the print data in the HD 226, the print data can be transferred to the data transferred via the network.
It can be received while printing, and processing that does not impose a load on the network can be executed. During the scanner processing, the image data read by the scanner unit of the copying machine 103 is stored in the buffer memory 209. This data is H after the area discrimination processing is executed.
It is stored in D226. Furthermore, when converting the page description language of the image data by the PDL conversion circuit 212, the HD 22
The image data stored in 6 is read and processed (page description language conversion of the image data will be described later). Therefore, it is not necessary to scan the image data again. Of course, the image data stored in the HD 226 can be used when the image data is needed again without being limited to the page description language. The HD 226 is connected to other blocks via the DISK I / F 225. The DISK I / F 225 manages the input / output of data of the HD 226.

[Various Hard Blocks] In order to realize the functions of the blocks described above or to realize the functions for which detailed description is omitted, there are the following blocks.

CPU 201: Executes a server program stored in the PROM 202 to deliver image data, control the scanner / printer server 102, and manage the copying machine 103.

PROM 202: Stores a server program and the like.

RAM 203: Used as a working memory of the CPU 201, temporary storage of image data, and the like.

Serial I / F 204: Used for connecting to a terminal or a serial printer via the serial I / F port 217.

Parallel I / F 205: Centronics I / F, GPI via parallel I / F port 218
This parallel I / F port 2 is connected to a parallel input / output scanner or printer composed of B, SISI, etc.
Input / output of image data is also performed from 18.

Communication control device 207: Data which has been character coded by character recognition from character image data is digital public line or IS via public line I / F 220.
It connects to DN and controls transfer or reception. However, although the communication control device 207 may be provided in the host computer, the communication control device 207 is provided in the server 102 here.

<Configuration of Copying Machine (FIGS. 3, 4, and 5)> FIG. 3 is a configuration diagram of the copying machine 103. CPU 306 is memory 3
The program stored in 05 is executed to control each block which is a constituent element. In this configuration, the copying machine 10
The operation of each part when the scanner side and the printer side operate independently of the operation No. 3 will be described. In the figure, a reading sensor 301 is a sensor for reading an image, and a scanner-side image processing circuit 302 is a circuit for processing an image signal read by the sensor 301. The printer-side image processing circuit 303 is a circuit that processes image data output by the printer engine 304, and the printer engine 304 records the image data on a medium. The scanner printer control CPU 306 controls the entire copying machine. Main memory 3
In 05, programs and data executed by the CPU 306 are stored. Each block that makes up the copier is
The scanner printer is connected by the image data bus 309. The sensor 301 and the image processing circuit 302 are collectively referred to as a scanner unit 310, and the printer engine 304 and the image processing circuit 303 are collectively referred to as a printer unit 311. Further, hereinafter, the scanner unit 310 or the printer unit 31
When described as 1, the resources shared by the scanner unit and the printer unit, such as the CPU 306 and the timing control circuit 308, may be included.

[Scanner Operation] The operation for scanning an image will be described with reference to FIG.

The serial interface 307 receives a scan command from the scanner / printer server 102 and sends it to the CPU 306.

Next, the CPU 306 determines the image size to be read, the scan start position of the image, etc. based on the received scan command.

The CPU 306 controls a scanner drive circuit (not shown) in the scanner unit 310, and the sensor 301 reads out an image for each pixel column as shown in FIG. The read image data is subjected to necessary image processing by the scanner-side image processing circuit 302. After that, the timing control circuit 308 performs CLOCK (clock signal) 108 and VSYNC (vertical synchronization progress) 10 at the timing of FIG.
9, HSYNC (horizontal synchronization signal) 110, and image data DATA (image data signal) 111 synchronized with this
The image data obtained from the image data bus 309 is output as. The timing chart of FIG. 12 is an example showing reading of data for one page.

The scanner printer server 102 receives a signal HSYNC (horizontal synchronization signal) input from the copying machine 103.
110, VSYNC (vertical synchronization signal) 109, CLOC
The image data 111 is synchronized with the K (clock signal) 108.
Read. In the scanner printer server 102, the video I / F control circuit 216 has the copying machine 10
3 receives the image data and stores it in the buffer memory 209.

[Printer Operation] The operation for printing an image will be described with reference to FIG.

The serial interface 307 receives a print command from the scanner / printer server 102 and sends it to the CPU 306. Next, the scanner printer C
The PU 306 determines the image size, the print start position of the image, and the like according to the print command.

The CPU 306 has a printer engine 304.
Is controlled to record an image for each pixel column as shown in FIG. At this time, as shown in FIG. 12, the timing control circuit 308 causes the synchronization signal CLOCK (clock signal) 108,
VSYNC (vertical synchronization signal) 109, HSYNC (horizontal synchronization signal) 110, and image data 111 synchronized with this
Is received from the server 102, the printer-side image processing circuit 303 performs necessary processing, and then the printer engine 30
Output by 4.

[Command Operation] The serial interface 307 communicates commands between the server 102 and the copying machine 103 by serial transmission using SPCOM (scanner printer serial command signal) 107. This data is from the scanner printer server 102 to the copying machine 103.
For the scanner function of, a prescan command, a scan command, etc. are sent. Further, the copying machine 103 sends a copy command and status information such as an abnormal operation.

Similarly, a print command or the like is sent from the scanner printer server 102 for the printer function of the copying machine 103. In addition, the copy machine 103 runs out of paper,
Status information such as paper jams and malfunctions are sent.

[Image Processing on Scanner Side] FIG. 4 is a block diagram of the image processing circuit 302 on the scanner side. 4, the image processing circuit 302 includes a shading correction unit 401, a resolution conversion unit 402, a scanner color conversion unit 403, a scanner gamma conversion unit 404, and a level conversion unit 405. These are connected to a bus (not shown), and the CPU
From 306, the parameters of each unit can be changed.

The sensor 301 reads an RGB color image, and the data is input to the scanner side image processing section 302.

In the image processing unit 302, the shading correction unit 401 first performs shading correction on the input image data, and then the resolution conversion unit 402 converts the image reading resolution. This resolution is 4
00 dpi (dot per inch), 200 dpi, 100 dpi
You can select from i, etc. This resolution is designated by the CPU 306.

The scanner color conversion section 403 carries out color conversion of the image. For example, here, the RGB color data is YCr
It is converted into a color space such as Cb or converted into Lab. Of course, in the case of passing through, it remains RGB. When reading in monochrome, black-and-white conversion is performed based on the correlation of RGB data. The CPU 306 designates the scanner color conversion. Scanner gamma converter 4
Reference numeral 04 performs gamma conversion by a LUT (look-up table) on the image data such as YCrCb after color conversion. The CPU 306 designates this LUT. The level conversion unit 405 converts the number of effective bits of one pixel.
For example, the lower bits of the 8-bit YCrCb data of each color after gamma conversion are discarded, and the dynamic range is converted to Y for 6 bits and Cr and Cb for 4 bits respectively. This level is specified by CPU3
06 will do.

[Image Processing on Printer Side] FIG. 5 is a block diagram of the image processing circuit section 303 on the printer side. The following processing is performed on the image data input to the image processing circuit unit 303.

The printer color conversion unit 501 controls the RGB of the image.
Color conversion to. Here, when an image is input in a color space such as YCrCb, it is converted into RGB.

The printer gamma conversion unit 502 performs gamma conversion on the input image. That is, a conversion of R '= f (R) G' = f (G) B '= f (B) is performed. This conversion is performed by an LUT (lookup table). This LUT is set by the CPU 306
Will do.

The masking conversion unit 503 performs masking conversion on the input image. Where the masking transformation is

[0073]

[Equation 1] Linear transformation of, or

[0074]

[Equation 2] It is obtained by the secondary conversion of

This conversion is performed using the LUT or the gate array. The CPU 306 sets the parameters of the LUT or the gate array.

In the black generation / undercolor removal unit 504, C = 255-R ″ M = 255-G ″ Y = 255-B ″ Bk = α (min (C, Y, M)) (Equation 3) C As shown by '= C-Bk M' = M-Bk Y '= Y-Bk, RGB data is converted into CMK, and black (Bk) generation and undercolor removal are performed from this data. (Look-up table) or gate array The CPU 306 sets the parameters of this LUT or gate array.

In the binarizing unit 505, the printer engine 3
If 04 is a binary printer, the image is binarized. As the binarization method, a simple binarization method, a dither method, an error diffusion method, and three types are switched and used. Printer engine 30
If 4 is a multi-valued printer, the binarization unit 505 is not necessary. The binarization process is performed by the gate array. The CPU 306 sets the binarization method, the binarization threshold value, and the like.

The above is the configuration of the copying machine 103. next,
The components of the scanner printer server 102 will be described in detail.

<Description of Character Recognition (FIGS. 6 and 7)> FIG. 6 is a diagram showing an internal block diagram of the character recognition circuit, and FIG. 7 is a diagram showing character cutting and feature extraction for character recognition. . In FIG. 6, reference numeral 601 denotes a character recognition control DSP (Dig
ital Signal Processor), 602 is dual port RA
M and 603 are feature extraction circuits, 604 is an identification circuit (dictionary), and 605 is a local bus (for character recognition).

The character recognition procedure is described in the above [Character recognition circuit].
This is as described in the above section, which will be described with reference to FIGS. 6 and 7. .

As shown in FIG. 7A, the character segmentation is
This is a process of detecting character gaps in the character image data stored in the buffer memory 209 in the vertical and horizontal directions and cutting out a character block. The cut-out character block is written in the dual port RAM 602 by the main CPU 201. The DSP 601 reads the image data of the character block and puts it in the feature extraction circuit 603. Feature extraction circuit 603
As shown in FIG. 7B, the character image data is finely divided into blocks, and the direction of the line in the block is detected.
Extract character features. Although FIG. 7 shows nine blocks surrounded by an ellipse, the blocks are actually divided into smaller blocks, for example, 64 blocks or more.

The result of feature extraction is sent to the discrimination circuit 604, and a character similar to the extracted feature is selected.
At this time, some similar characters are selected and all the character codes are written in the dual port RAM 602.
From these candidate characters, the main CPU 201 detects the corresponding character based on the context of the character. The character is recognized as described above.

[Image Encoding / Decoding (FIG. 8)] FIG. 8 is a diagram showing internal functional blocks of the image encoding / decoding circuit 211. This block diagram is for performing ADCT encoding, which is a method of compressing an RGB multi-valued (8-bit for each color) image.

In this ADCT method, the R (Red) component input unit 801, the G (Green) component input unit 802, and the B component are input.
When RGB signals are input from the (Blue) component input unit 803, the color conversion unit 804 converts the signals into Y, Cr, and Cb (luminance and chrominance components). The converted signal is extracted for each component into a Y (luminance) component input unit 805, a Cr (chrominance) component input unit 806, and a Cb (chrominance) component input unit 807, from which the signal of each component is extracted. , 8 × 8 DCT (Discrete Cosine T
ransform) input to the processing unit 808. DCT processing unit 8
In 08, each signal is transformed by a two-dimensional discrete cosine transform (DCT) for each 8 × 8 block.

The signal after the DCT is converted into an alternating current (A
The AC component is divided into a C) component and a direct current (DC) component, and the AC component is subjected to a zigzag operation in a zigzag operation / image data input unit 812 and then two-dimensional Huffman encoded by a two-dimensional Huffman encoding unit 813. On the other hand, the direct current (DC) component is subjected to one-dimensional prediction by the one-dimensional prediction unit 810, and then Huffman coding is performed by the Huffman coding unit 811. These two code amounts are multiplexed by the multiplexing unit 814 to form a compressed image.

The decompression method can be obtained by expanding in the opposite direction of the ADCT coding block diagram of FIG. 8, that is, from the multiplexing section 814 side.

The color conversion unit 804 converts RGB data into YCrCb, and is unnecessary if the scanner unit has a function like the scanner color conversion unit 403 shown in FIG. You can set it to the through mode. Further, FIG. 8 shows a method of encoding / decoding color multi-valued image data, but other methods are sufficiently conceivable. For example, MH, MR, MMR and LZ used in FAX
Methods such as W (Lempel-Ziv & Welch) can also be configured.

<PDL Conversion Circuit (FIG. 9)> FIG. 9 is a partial circuit configuration of the PDL conversion circuit 212 for converting the image data read by the scanner into data described in PDL (page description language). Indicates.

The image data is read by the scanner section of the copying machine 103 and temporarily stored in the buffer memory 209. The CPU 901 analyzes the input image data, determines the type of image included in the input image data, and divides the region for each type. The divisions are characters, figures, and images, and their respective areas and position information are recognized. The data obtained during the area discrimination process, for example, the data such as labeling is stored in the RAM 902 and the process proceeds. The CPU 901 uses the information obtained by the area discrimination to recognize the character from the image data by the character recognizing means to encode the character portion, and to form the page description language together with other information. The image and graphic areas are made into a page description language together with position data of image data. The character recognition means may be provided in the PDL conversion circuit by the same procedure as that described for the character recognition circuit 210, or may be realized by using the character recognition circuit 210. .

These processes are shown in the flow charts of FIGS. The control program of this flowchart is stored in the ROM 903 and executed under the control of the CPU 901.

The processing of FIG. 29 can be roughly divided into four stages. The four stages are the initial setting process (steps S2501 and S2502) and the process of determining the area of the image information (steps S2503 to S250).
2509), a process of discriminating detailed information in the image information (steps S2510 to S2518), and a process of converting the image into data described in PDL (steps S2519 to S2521).

Now, each step of FIG. 29 will be described in detail.

In step S2501, a command indicating the type of PDL is input from the host computer 101 via the network 104. In the subsequent processing, the image information is described in PDL of the type designated at this time. Then, in step S2502, a program is selected according to the type of PDL in order to extract information suitable for the designated type of PDL from the image information according to the input command.

Here, FIG. 26 shows CaPSL and PostS.
Two types of PDL called "script" are shown as an example. In accordance with this, the difference in code depending on the type of PDL is shown.

FIG. 26 shows an example of characters and separators using the character string in the left column.

The following three pieces of information are required to convert the character part into the page description language.

1. Start position information 2. Font information (character size, character pitch, character style) 3. The position and size of the character area of the character code information image are determined by the area determining means. The character code is recognized by the character recognition means. The character size and character pitch can be known from the character width and character pitch determined by the area discriminating means. The character style (font type) can be recognized if the character recognition means is provided with a function for judging the characteristics of the character, but here it is fixed to a predetermined font for simplicity. On the host computer side, the received information is P
With the DL format, it is easy to change the character style. Therefore, it doesn't matter if the characters are fixed to a particular style.

In order to convert the separator part into the page description language, the following three pieces of information are necessary.

1. Start position information 2. Line information, line start information 3. Line end information FIG. 27 shows an example of image data such as a natural image. In the case of a natural image or the like, image data is stored in the form of a bitmap or a compressed form thereof, and therefore the following three pieces of information are required to convert the natural image into the page description language.

1. Start position information 2. Raster image drawing information (total number of bytes, width, height, resolution, compression format) 3. The image data start position information is recognized by the area discriminating means. The resolution specified when the scan command is issued by the host computer is sent by a command. Therefore, the resolution can be obtained by holding this command. The width and height are obtained from the coordinate values of this rectangle by the region discrimination means discriminating the rectangle around the image data. The total number of bytes is calculated from the width, height and resolution. As a compression format, a command for designating an image coding method is input from the host computer side when a scan command is issued, and image compression is performed in the format designated by using the image coding circuit 211 in the server 202. The compression format can be obtained by holding the specified command at this time.

FIG. 28 shows an example of a graphic image. The area where the figure is drawn can be determined by checking the pixel density in the rectangle recognized as the image area. However, although it is possible to recognize and encode a figure with a simple figure, it is very difficult to recognize a figure with a line drawing or the like. Therefore, the graphic area is treated as image data like the area of the natural image. In this case, the following three pieces of information are required to convert the graphic part into the page description language.

1. Start position information 2. Raster image drawing information (total number of bytes, width, height, resolution, compression method) 3. Image data These data are obtained by the same method as in the case of the natural image described above.

As described above, the data required for PDL conversion differs depending on the nature of the image, and the data is converted into different codes as shown in FIGS. 26, 27 and 28 depending on the type of PDL. Therefore, step S250
In order to convert the image data into the PDL designated in No. 1, processing is performed by the program corresponding to the designated PDL.

Returning to the explanation of FIG.

When the initialization process is completed, step S2
In 503, the image information of the PDL conversion target designated by the host computer 101 is input from the scanner unit of the copying machine 103. The input data is the step S
At 2504, the data is accumulated in the buffer memory 209 of the server 102, and the area discrimination processing, which is the preprocessing of character recognition and PDL conversion, is started under the control of the CPU 901.

The area discriminating process is performed on the original read by the scanner, which may include characters, figures, and natural images. After the area discrimination processing, character recognition processing and PD
When converting to L, the image data to be processed is required again. Therefore, the following method is conceivable as a method of securing the image data necessary after the area determination.

1) The capacity of the buffer memory 209 is increased so that one page can be stored.

2) The buffer memory 209 serves as a band memory, and is a magnetic disk or the like that can be temporarily stored in the server 20.
Hold within 2.

3) The buffer memory 209 serves as a hand memory and performs image input for area determination and image input after area determination by scanning the original document.

Although any of the above methods may be used, in this embodiment, the server has the HD 226, and the method (2) can be adopted. Hereinafter, it is assumed that the server 202 has a means for re-obtaining the image data that is the object of the discrimination after the region discrimination.

FIG. 21 shows an example of an image original containing characters and natural images. In the figure, the character portion is divided into some sections. The image in FIG. 21 is set as a target for future processing, and this image is read from the scanner and read from the buffer memory 20.
9 is stored.

In step S2505, the image data accumulated in the buffer memory 209 is thinned out using a matrix of m * n to reduce the data. In the thinning-out process, as shown in FIG. 22A, when there is even one dot of data of a predetermined threshold value or more in the m * n matrix, the dot information corresponding to the matrix is set to "1". Further, as shown in FIG. 22B, when there is no data equal to or larger than a predetermined threshold value in the matrix, the dot information corresponding to the matrix is represented by "0". The data obtained by this thinning-out process is saved in the work area of the RAM 203.

The thinning-out process is a means for speeding up the region discrimination, and the thinned-out information is sufficient for discriminating the characteristics of the image information.

In step S2506, labeling processing is performed. The labeling process is a process for searching for pixels of “1” (hereinafter referred to as black pixels) line by line and adding a label to the image information obtained by the thinning process.
In addition, the same label is attached to pixels that are consecutive in the vertical, horizontal, and diagonal directions, and at the same time, a rectangular shape is used to group the pixels. The label here is for distinguishing consecutive pixels, and in the labeling process, the pixels are
Replace with characters such as B and C or character strings.

A labeling processing procedure will be described with reference to FIG. The image information illustrated in FIG. 23A has a black pixel at the third pixel from the left in the first row. Therefore, the label "A" is added to the third pixel. The white pixel portion is "0". There is a black pixel on the fifth pixel, but the fourth pixel is a white pixel and is not continuous with the black pixel on the third pixel, so a label "B" different from the third pixel is added to the fifth pixel. 1st row of 8
Black pixels from the 10th pixel to the 10th pixel are next to each other,
Since it is not continuous with "B" of the fifth pixel, a label of "C" is added to each. Next, the second line is searched. Two
The third pixel in the row has a black pixel, and the pixel that is continuous with this black pixel in the upward direction is already labeled as "A". Therefore, this black pixel is also labeled with the same label "A". The fourth pixel is labeled "A" because it is the black pixel "A" on the left. Here, the 5th pixel in the first row was labeled as "B", but since it is diagonally continuous with the 4th pixel in the second row, the label of the 5th pixel in the first row is changed to "A".

By repeating the same processing, as shown in FIG.
The image of (1) can be labeled with the label A and the label C as shown in FIG. 23 (2).

When the labeling process is completed in this way,
In step S2507, rectangular processing is performed. This process is a process of enclosing continuous black pixels in a rectangular frame. For this reason, the maximum vertical and horizontal tangent lines of each label are stored. The maximum circumscribed line is obtained by the maximum value and the minimum value of (x, y) of blocks of pixels having the same label. As a result, a region including continuous black pixels becomes a rectangle with one label. The labeled data is stored in the work area of the RAM 203 in step S2508.

The above thinning processing and labeling processing are
This is performed every time image data is input to the buffer memory 209, and it is determined in step S2509 whether all the images have been finished.
Repeat until finished.

After the labeling process (step S2506) and the rectangle process (step S2507), the rectangle width W, height H, area S, the number of pixels per area, that is, the rectangle size such as the pixel density D is extracted from the extracted rectangle information. Information is extracted, and using this rectangle size information, characters, figures, natural images,
In step S2510, the attributes of a rectangle such as a separator that separates paragraphs of a sentence by a table, a frame or a line in the sentence are determined.

The criterion for determining the rectangular size information for discriminating each attribute will be described.

As shown in FIGS. 24 (1) and 24 (2), a rectangle in which a character exists has a width W and a height H of the rectangle that are continuous in a predetermined size, and a proper space exists in the rectangle. , Character area.

As shown in FIG. 25 (3), a rectangle in which a figure exists has a size of the rectangle above a certain level, and the pixel density D in the enclosed area is below a certain value. To judge.

As shown in FIG. 24 (4), a rectangle in which a natural image exists has a size equal to or larger than a predetermined size, and the pixel density D in the enclosed area is equal to or larger than a predetermined value. Judge as an area.

As shown in FIG. 24 (5), the rectangle in which the separator exists has a width W or height H of the rectangle that is equal to or longer than a certain length (when the rectangle is continuous). Has a certain width W and height H and is continuous (in the case of discontinuity as indicated by a broken line), or the width W and height H of the rectangle are not less than a certain size, If there is a similar circumscribed line on the inside and outside (in the case of an enclosing line), etc., it is judged as the separator area.

As shown in FIG. 24 (6), the rectangle in which the table exists has a rectangle larger than a predetermined size, and the pixel density in the enclosed area is below a certain value (that is, there is a blank inside). In the case of, and is separated by a line such as a separator, it is determined to be a table area.

The determined attribute information is stored in the RAM 203 together with the rectangular information.

When the attributes of all rectangles of the input image are determined in step S2510, more detailed feature determination is performed for each attribute.

For that purpose, first, in step S2511,
The rectangle information stored as a separator in the RAM 203 is read, the line thickness and the line type (solid line or broken line, etc.) are determined from the rectangle size information, coded, and coded in the RAM 20.
Save to 3. Further, the rectangular information stored as a table in the RAM 203 is read, and the position information and the size information of the frames forming the table are encoded. The character area inside the table is newly registered in the RAM 203 as a character rectangle, and the characters are recognized and coded by the character recognition function described later.

Further, in step S2512, the RAM 203
The rectangular information stored as a natural image / figure is read and the position and frame information is encoded. The process of recognizing a figure and coding it is not performed because it is quite complicated.
Here, the graphic region is used as the attribute of the region, and the image data is treated as a bitmap like the natural image. However, when performing the compression process, a process suitable for the type of image data is performed.

In step S2513, RAM2
The rectangular information stored in 03 as a character area is read out, a histogram of black pixels is calculated at several points in each of the vertical and horizontal directions, and the characteristics thereof, that is, the projective analysis is performed. FIG. 25 shows how the histogram is taken. In step S2514, it is determined from the variance of the amount of black pixels and the interval of white pixels whether the character string in the rectangular area is vertical writing or horizontal writing.

In step S2515, the height H and width W of the read character rectangle information is averaged in accordance with the set direction determined in step S2514 to determine the character width.

In step S2516, the space between rectangles is obtained from the read character rectangle information, and the character pitch is analyzed.

In step S2517, step S25
Characters that do not match the character width determined by 16 are determined as heading characters. Also in the headline character area, the height H, width W, character size, and pitch of the rectangle are analyzed and coded.

In step S2518, the code information obtained by the analysis in steps S2511-S2517 is stored in the RAM 203 together with the attribute information.

When the information analysis for each rectangle by the area discriminating means is completed as described above, step S2519
Then, the image data is read again in order from the first position, and stored again in the buffer memory 209 in step S2520. Then, from step S2521, the process of converting the image data accumulated in the buffer memory 209 into PDL is started.

In step S2521, one band of image data accumulated in the buffer memory 209,
The PD specified in step S2501 based on the image attribute data and position data stored in the RAM 203.
Convert to L. The detailed procedure of the processing in step S2521 is shown in the flowcharts of FIGS. 31 to 32 and will be described individually.

The flowchart of FIG. 31 shows the processing of the character portion and the in-table character portion.

First, in step S2601, it is determined whether the image data stored in the buffer memory 209 in step S2520 has a character portion or a character portion in the table according to the attribute information and the rectangular position information stored in the RAM 203. To do. If the character exists, the processing of this character portion or the character portion in the table is continued, and if the character does not exist, the processing branches to the symbol E of FIG. 32 and shifts to the processing of natural images and graphics.

If the character portion exists in the buffer, it is determined in step S2602 from the position information in the RAM 203 whether or not all the character portion is stored in the buffer memory 209.

If it is determined that the data has not been received, the process branches to step S2606 to perform joint processing. In step S2607, it is determined whether or not the joint is the upper portion. If not, the process branches to step S2609 to store the joint image data in the RAM 203.

If it is determined that the joint is the upper portion, in step S2608, the joint image data of the upper band already stored in the RAM 203 is added to the image data stored in the buffer memory, and step S2603 is performed. Branch to.

On the other hand, if it is determined that the character portion accumulated in the buffer memory 209 is stored in the memory, or if the addition of the joint data in step S2608 is completed, it is included in the image data in step S2603. The recognized character is recognized by the character recognition circuit 210 and converted into a character code. And step S
2604, the RA obtained by the area discrimination processing
A RAM that stores the rectangular position information that had the character information stored in M203 and the obtained character code together.
Save again to 203.

The above processing is repeated while checking in step S2605 whether the processing of the character area in the buffer memory 209 is completed.

The flowchart of FIG. 32 shows the processing of the natural image portion and the graphic portion.

First, in step S2701, whether or not there is a natural image portion or a graphic portion in the image data stored in the buffer memory 209 in step S2520 is determined according to the attribute information and the rectangular position information stored in the RAM 203. to decide. If there is a natural image or graphic, the processing of the natural image portion / graphic portion is continued, and if it does not exist, the process branches to step S2801 for PDL conversion processing.

If natural image or graphic data exists in step S2701, it is determined in step S2802 whether a compression instruction command has been input from the host computer for these data. Note that the compression instruction command is input together with the command indicating the type of PDL in step S2501.

If there is a compression instruction, step S27.
Go to 03. In step S2703, the image data of each area is compressed and coded by a designated compression method such as ADCT processing for a natural image and MMR processing for a graphic.

When the compression is completed, in step S2704, the compression code is re-stored in the RAM 203 together with the position information and the like obtained by the area discrimination processing and stored in the RAM 203. However, if this natural image or figure is continuous with the previous buffer, it is saved continuously with the data in the previous buffer.

On the other hand, if it is determined in step S2702 that there is no compression instruction, the image data of the graphic area and the natural image area accumulated in the buffer memory 209 is stored in the RAM 203 together with the position information and the like stored in the RAM 203.
Save it in AM203 as it is.

When the storage of the data in the buffer memory 209 is completed, it is judged in step S2705 whether or not the processing is completed for all the target images. If it is completed, the processing for converting to PDL in step S2801 is executed. And move on. If it is determined that the processing is not completed, step S2
Returning to 519, processing of the next band is started.

When the area discrimination of the image data, the attribute discrimination, and the processing of the code suitable for each attribute are completed, step S2
Proceeding to step 801, as shown in FIG. 26 and described above,
The image data is converted into the designated PDL and is stored in the RAM 203 as the image data converted into the PDL.

The image data can be converted into PDL as described above.

<Description of Automatic Translation (FIG. 10)> FIG. 10 is a circuit configuration diagram for realizing an automatic translation circuit 213 for performing Japanese-English / English-Japanese translation.

The character-coded characters are written in a dual port RAM 1007, which is connected to the main CPU bus 223 and the local bus 1008 and is accessible from both as a group for each clause. Translation circuit 21
On the third side, the CPU 1001 reads out the phrase and performs a morphological analysis. The words of each sentence are based on the registered words in the word dictionary ROM 1004. Next, after performing a hooking analysis, a semantic analysis is performed. Also at this time, the word dictionary ROM 10
Based on 04. Next, the words in the sentence are rearranged by the replacement rule analysis based on the replacement rule dictionary 1005. After rearranging sentences, English / Japanese
Based on the Japanese-English dictionary 1006, a Japanese-to-English translation or an English-to-Japanese translation is performed, written again in the dual port RAM, and read out by the main CPU 201. The ROM 1003 is a CP for realizing the above processing.
It stores the programs that U1001 executes,
The RAM 1002 is a memory used at that time.

<Operation of Scanner / Printer> Next, the operation of the character recognition scanner / printer server of the present embodiment having the above-mentioned configuration will be described. In the description, the arrows represent sequential operations.

[Scanner] (1) Prescan-> A read instruction is given by the client process 105 on the host computer side.

→ The instruction is accepted by the server process 106 on the scanner printer server side.

→ An image is read by the scanner unit 310 of the copying machine 103.

→ The server 102 buffers the read image data in the memory 209.

→ The thinned data or the compressed image data is transferred to the host computer 101 through the network 104.

→ The host computer side client process 105 displays the thinned data or the data obtained by expanding the compressed image data on a monitor (not shown).

The image read by the copying machine 103 is received by the host 101. (2) Character recognition scan → For the image data obtained in prescan (1),
After the character area is designated, the client process 105 instructs the image reading again.

-> The reading instruction is accepted by the scanner printer server side server process 106.

→ The area designated by the scanner unit 310 of the copying machine 103 is read.

→ The server 102 buffers the read image data in the memory 209.

→ Character cutting is performed on the character area of the buffered image data.

The character recognition circuit 210 encodes characters.

→ The character code is stored in the memory 203.

→ The character code is transferred to the host computer 101 through the network 104.

As described above, the image is read by the copying machine 103,
It is recognized as a character and received by the host 101.

The following processing may be added to the character recognition scanning operation.

When extracting a character from a buffered character image, the character color or background color is detected, and the character of the designated color or a color other than the designated color is recognized, and the character code is transferred to the host computer. To do.

The position of the character is detected by extracting the character from the buffered character image, the position of the character is added to the character code, and the character code is transferred to the host computer 101.

The position of the character is detected by extracting the character from the buffered character image, and the obtained character code and position information are converted into a page description language form and transferred to the host computer 101.

After detecting the font type and size of the character code for which character recognition has been performed, the font type and size are changed according to a prior instruction from the host computer 101.

The read character image is held in the server 102 in the form of the page description language, and the image data is again combined with the image data (natural image or the like) obtained by scanning and transferred to the host computer 101.

The read character image is held in the server 102 in the form of page description language, and the image data obtained by scanning the image data again is compressed and synthesized, and transferred to the host computer 101.

The graphic data and the image data are also converted into the page description language and transferred to the host computer 101 together with the character data. However, as described regarding the processing by the PDL conversion unit, it is not necessary to specify a region such as a character region, a graphic region, or an image region. The recognition of the area for each image type can be realized by the area determination procedure as described above. Further, even when only the characters are recognized, the character area can be recognized, and therefore the host need not specify.

The above operation can also be performed. on the other hand,
The server operates as follows for the printer of the copying machine 103.

[Printer] (3) Print → Transfer print data from the host computer side client process 105 to the server 102.

-> The scanner printer server side server process 106 receives print data.

If the print data is compressed image data,
The encoding / decoding circuit 211 performs bit expansion.

If the print data is character code data, character bit expansion is performed.

If the print data is bit data, it is left as it is and no particular processing is performed.

→ The bit data is accumulated in the buffer 209 in the server 102.

→ The data in the buffer memory 209 is sent to the copying machine 103 and printed by the printer unit 311.

When the client is used as a character recognition server by using the character recognition function in the scanner printer server 102, the following operation is performed. (4) Character recognition server → The character image data is transferred from the client computer side 105 of the host computer to the server 102.

→ The scanner printer server side server process 106 receives character image data.

→ If the character image data is compressed image data, the encoding / decoding circuit 211 performs bit expansion.
If the character image data is bit data, no particular processing is performed as it is.

→ The character image data is stored in the buffer memory 209.
Accumulate temporarily in.

→ Character extraction is performed on the buffered character image data.

The character recognition circuit 210 encodes character image data.

→ The character code is stored in the memory 203.

→ The character code is transferred again to the host computer 101 through the network 104.

With the above operation, the server functions as a character recognition server that processes the data given by the host 101 and returns it to the host.

Further, the following procedure can be added as the operation of the character recognition server.

At the time of extracting the character of the buffered character image, the character color or the background color is detected, the character of the designated color or the color other than the designated color is recognized, the character code is formed, and the character code is transferred to the host computer 101. The position of the character is detected by the character cut-out of the buffered character image and added to the character code, and the character code is transferred to the host computer 101. The position of the character is detected by the character cut-out of the buffered character image. The obtained character code and position information are converted into a page description language form and transferred to the host computer 101. Further, by utilizing the character recognition function in the scanner / printer server 102, the copying machine 103 is subjected to character recognition. The operation when used as a copy is as follows. (5) Character recognition copy → A character original is read by the scanner unit 310.

→ The server 102 stores the image data in the memory 20.
Buffering to 9.

The character recognition circuit 210 encodes the buffered character image. At the same time, identify the font type and size.

→ The character code, font type and size are stored in the memory 203.

→ The font and size of the character code are changed to those designated by the host computer 101 in advance.

→ Perform bit expansion of character code data.

→ The bit data is accumulated in the buffer 209 in the server 102.

→ The data in the buffer memory 209 is sent to the printer unit 311 of the copying machine 103 and printed.

The following operation is performed when used as a character recognition / automatic translation server that automatically translates a character string and a sentence that are character coded by character recognition. (6) Character recognition / automatic translation server-> A character region is designated for the image data obtained in the prescan (1), and the client process 105 again gives an instruction to read the image to the server 102.

→ The instruction is accepted by the server process 106 on the scanner printer server side.

→ The designated area is read by the scanner section 310 of the copying machine 103.

→ The server 102 stores the image data in the memory 2
Buffer to 09.

-> Character extraction is performed on the buffered character image data.

The character recognition circuit 210 encodes characters.

→ The character code is stored in the memory 203.

→ The character code is transferred to the host computer 101 through the network 104.

→ The client process 105 receives the character code and automatically translates it based on the character string and the preceding and following sentences.

For example, translation is performed between English to Japanese, Japanese to English, and other two languages.

The above is an example in which the automatic translation is performed by the host computer, but the automatic translation circuit 213 in the server 102 is used.
The translated character code may be sent to the host computer 101 through the network 104 after being translated by using the translation function using.

Further, the operation when used as a character recognition / teletext communication server for transmitting or receiving a character-recognized and character-coded document using a public line is as follows. (7) Character recognition / teletext communication server transmission-> A character area is designated for the image data obtained in the prescan (1), and the client process 105 instructs the server 102 to read the image again.

→ The printer process is accepted by the scanner printer server side server process 106.

-> The designated area is read by the scanner section 310 of the copying machine 103.

→ The server 102 stores the image data in the memory 2
Buffer to 09.

→ Character extraction is performed on the buffered character image data.

The character recognition circuit 210 encodes character image data.

→ The character code is stored in the memory 203.

→ The character code is transferred to the host computer 101 via the network 104.

→ The client process 105 receives the character code and transfers it to the other party through a public line connected to the host computer 101, for example, a digital public line or ISDN. Reception → The data transferred from the sender is received by the host computer 101 and transferred to the scanner / printer server 102 through the network 104.

→ The scanner printer server side server process 106 receives the data.

→ If the data is compressed image data, the image encoding / decoding circuit 211 performs bit expansion.

If the data is character code data, character bit expansion is performed.

If the data is bit data, no particular processing is performed as it is.

→ The bit data is accumulated in the buffer 209 in the server 102.

→ The data in the buffer memory 209 is sent to the copying machine 103 and printed by the printer unit 311.

In the above-mentioned transmission and reception, an example in which the public line function is performed by the host computer has been shown, but the server 102 is connected to the public line with the communication control device 207 having the public line connection function and converted from the character image data. The character code may be sent to the other party via the public line. On the contrary, the server 102 receives the data transferred from the public line by the communication control device 207, and the copier 1
It is also possible to print from the printer unit 311 of 03.

<Explanation of Image Input / Output Procedure> Next, the image input / output of the scanner / printer server 102 of the present embodiment will be described with reference to the zigzag chart while following the exchange of messages.

When the image is read from the copying machine 103, the image size, the image position, the resolution, the format (dot sequential, line sequential, plane sequential), edge enhancement, color space (R
GB or YCrCb), color (which color to send, for example only G), level (the number of gradations of color), coding method (ADCT or not coded), bit rate (bit rate at the time of coding), It is necessary to specify parameters such as thinning rate during prescan, which file to read, character recognition setting (whether to do it, color specification, page description verbalization, etc.).

Therefore, the client process 105
The user is instructed to specify these parameters via a user interface such as a display. The user receives the instruction, specifies the parameter, and instructs the host computer 101 to execute the prescan. Then, the client process 105 communicates with the server process 106 according to the sequence of FIG.

Before describing FIG. 13, the commands exchanged between processes and the commands to the copying machine will be described with reference to FIGS. 16, 17, 18, and 19.

<Packet structure (FIG. 16, FIG. 17, FIG. 1)
8. FIG. 19)> The configuration of packets exchanged between the host computer 101 and the scanner printer server 102 will be described with reference to the drawings. However, the explanation centered on what appeared in the zigzag chart described later.

16 and 17 show the packet structure.
Each field that constitutes a packet will be described.

FIG. 16A shows a PRESCAN packet,
16B is a SCAN packet, and FIG. 16C is a PR.
It is an INT packet, and the first byte of each packet is a tag that indicates what the packet is. For example, P
If it is RESCAN, it indicates that pre-scan is performed by "1", and if it is SCAN, it indicates that scanning is performed by "2".

XSIZE is the X of a 2-byte image.
Indicates the size of the direction.

YSIZE is the Y of an image consisting of 2 bytes.
Indicates the size of the direction.

XSTART indicates a scan / print start position of an image composed of 2 bytes in the X direction.

YSTART indicates the scanning / printing start position of the 2-byte image in the Y direction.

XZOOM is the X of a 1-byte image.
Indicates the scan and print resolution in the direction.

YZOOM is the Y of an image consisting of 1 byte.
Indicates the scan and print resolution in the direction.

FORMAT indicates a scanning method of an image,
Specify 1 for dot sequential, 2 for line sequential, and 3 for frame sequential.

EDGE / SMOOTH is edge enhancement,
The degree of smoothing is shown, 16 to 1 are edge enhancement, −
If it is 1 to -16, it is designated as smoothing.

COLORTYPE indicates the color space of the image, 1 for RGB, 2 for GMY, 3 for CMYK.
Specify 4 for CrCb, 5 for Lab, and 6 for XYZ. In the case of RGB, the first color will be called R, the second color will be called G, and the third color will be called B. In the case of YCrCb, the first color will be Y, the second color will be Cr, and the third color will be called. I will call it Cb.

COLOR indicates which color of the image colors is to be sent. Which color is assigned corresponding to the bit. For example, the 0th bit is set for the first color only, the first bit is set for the second color only, the second bit is set for the third color only, and the third bit is set for the fourth color.

For example, COLORTYPE is RGB,
When sending all RGB colors, the number is 7 (= 00000111B). However, "B" attached to the numeral indicates that it is a binary number.

If two colors of R and B are sent, 5
(1st color = 1, 3rd color = 4; 00000101B). Similarly, when COLORTYPE is YCrCb and only Y is sent, it becomes 1.

LEVEL consists of 2 bytes, the first 4 bits indicate the number of gradations of the first color, and the next 4 bits are the second.
The gradation number of the color is shown, and the next 4 bits show the gradation number of the third color. The last 4 bits are undefined.

The number of gradations represented by these bits is designated by an index of 2, such as 256 gradations if 8 and 4 gradations if 6.

CODE indicates an encoding method, 0 when not encoding, 1 when MCT encoding, MMR
If it is 2, then specify it as 3 if it is LZW.

BITRATE indicates a coding rate of coding, 6 indicates 1/6, and 12 indicates coding at a compression rate of 1/12.

XSTEP and YSTEP specify to what extent images are thinned out and sent at the time of prescan. For example, if you want to send images every 5 pixels vertically and horizontally, use XSTEP
= 5, YSTEP = 5. When thinning is not performed, XSTEP = 0 and YSTEP = 0.

OCR indicates whether character recognition is performed or not. Set 1 when character recognition is performed on the data in the scan operation area, and set 0 when character recognition is not performed. Further, 2 is set when the position information is added to the character recognition function, and 3 is set when the page description language is used. Further, when a character of a designated color is detected when recognizing a character, and when recognizing only that character, 4 is set and the character of the color designated by COLOR is recognized. When characters are recognized only in the background color area of the designated color, 5 is set and characters written in other colors are recognized in the area designated by COLOR. In addition, the relationship between colors and characters can be expanded as an option.

PDL is data for designating whether or not page description language (PDL) is applied to scan data. Set 0 when not using PDL. PD
In the case of L conversion, a predetermined value is set for each type of PDL. For example, PDL called CaPSL
If 1 is specified, 1 is set, and if a PDL called PostScript is specified, 2 is set.

The above is the PRESCAN packet and SCA.
It is a description of each field of N packets. Although common parts with these are omitted, each field of the PRINT packet will be described.

UCR178 indicates α at the time of black generation (see (Equation 3)).

BI-LEVEL 179 indicates a binarization method, and 0 indicates that binarization is not performed when the printer engine 304 is a multi-value printer. Reference numeral 1 is a fatting pattern of the dither method, 2 is a Bayer pattern of the dither method, 3 is a simple binarization method, and 4 is an error conversion method.

THRESHOLD 180 indicates the binarization threshold value (0 to 255) of the simple binarization method.

PAGE 181 indicates the number of pages at the time of printing.

FIG. 17D shows an OK packet. OK
The packet is only a 1-byte OK tag.

FIG. 17E is a GAMMA packet, and the S / P 182 of the second byte indicates which gamma table of the scanner image processing circuit 302 or the printer image processing circuit 303 of the copying machine 103 is set. . After that, the gamma table is composed of 768 bytes for 256 * 3 colors.

FIG. 17 (f) shows a MASKING packet, which is composed of a masking tag and a masking parameter aij consisting of a fixed point of 2 bytes. aij is i row j
The components of the matrix located in columns.

FIG. 17 (g) shows a STATUS packet.
It is composed of a status tag, the number N183 of statuses, and STATUSn (n = 0 to N) indicating the status.

FIG. 17 (h) shows a DATA packet, which includes a data tag, the number N of the following image data and the image data 1, 2 ,.
3 ... and.

FIG. 16 (i) shows an ESC packet, which is a sequence in which the first byte starts with ESC, and is expanded by the interpreter 212 into a bit map or byte map image. This is equivalent to a normal printer ESC sequence.

[Composition of command] The structure of commands such as prescan command, scan command, status information, gamma setting command, masking setting command, etc. exchanged between the server 102 and the copying machine 103 is also shown in FIGS. Communication is performed in a similar format.

18 and 19 show the server 102 and the copying machine 1.
The structure of the instruction during the period 03 is shown. 18A shows a prescan instruction, FIG. 18B shows a scan instruction, and FIG.
Is a print instruction. The first byte of each packet is a tag that describes what that packet is. For example, P
If the packet is a RESCAN packet, the tag is 1, indicating that prescanning is performed. In case of SCAN, 2 indicates scanning.

Next, each field of the packet in each packet of prescan, scan and print will be described.

XSIZE and YSIZE are 2 bytes and indicate the size of the image in the X direction and the size of the image in the Y direction, respectively.

XSTART and YSTART are 2 bytes and indicate the X-direction scan and print start position of the image and the Y-direction scan and print start position of the image, respectively.

XZOOM and YZOOM are 1 byte and indicate the X-direction scan and print resolution of the image and the Y-direction scan and print resolution of the image, respectively.

FORMAT indicates a method for scanning an image,
Specify 1 for dot sequential, 2 for line sequential, and 3 for frame sequential.

EDGE indicates the degree of edge enhancement and smoothing. 16 to 1 designates edge enhancement, and -1 to -16 designates smoothing.

COLOR TYPE indicates the color space of the image, 1 for RGB, 2 for CMY, 3 for CMYK.
Specify 4 for YCrCb, 5 for Lab, and 6 for XYZ. In the case of RGB, the first color is called R, the second color is called G, and the third color is called B.
In the case of b, the first color is Y, the second color is Cr, and the third color is C.
I will call it b.

COLOR indicates which one of the colors of the image is to be sent. Bits are assigned to which color.
For example, the 0th bit is set for the first color only, the first bit is set for the second color only, the second bit is set for the third color only, and the third bit is set for the fourth color.

For example, when the COLOR TYPE is RGB and sends all RGB colors, the 0th, 1st and 2nd bits are set to 1 to be "7" (00000111).

When sending two colors of R and B, "5" (first color = 1, third color = 4: 0000010) is similarly performed.
It becomes 1). Similarly, COLOR TYPE is YCrC
In case of sending only Y in b, it becomes "1".

LEVEL consists of 2 bytes, the first 4 bits indicate the gradation number of the first color, and the next 4 bits are the second.
The gradation number of the color is shown, and the next 4 bits shows the gradation number of the third color. The last 4 bits are undefined. The value of the LEVEL field is the number of gradations of each color.
If so, it is designated by an index of 2 like 4 gradations.

CODE indicates a coding method, and is designated as 0 when not coded and as 1 when coded by ADCT.

UCR represents α at the time of black generation.

BI-LIVEL indicates a binarization method, and 0
Indicates that the print engine 606 is a multi-value printer and binarization is not performed. Reference numeral 1 is a dithering fatting pattern, 2 is a dithering Bayer pattern, 3 is a simple binarization method, and 4 is an error diffusion method.

THRESHOLD indicates the binarization threshold value (0 to 255) of the simple binarization method.

PAGE indicates the number of pages at the time of printing.

FIG. 18D is one type of status information.
Indicates OK. OK is a 1-byte OK tag.

FIG. 19E shows a gamma setting command. The S / P of the second byte indicates which gamma table of the scanner image processing circuit 302 or the printer image processing circuit 303 of the copying machine 103 is to be set. . After that, the gamma table is composed of 768 bytes for 256 * 3 colors.

FIG. 19 (f) shows a masking setting command, which is composed of a masking tag and a masking parameter consisting of a 2-byte fixed point. The contents have the same meaning as in FIG.

FIG. 19G shows status information, which is made up of a status tag, the number N of statuses, and N statuses (STATUS 1 to N).

FIG. 19H shows a copy instruction, which is a 1-byte copy tag only.

Next, a command sequence for using the copying machine 103 from the client process 105 will be described.

<Description of Command Sequence (FIGS. 13 to 1)
5)> [Scanner Side] In FIG. 13, the client process 105 first determines the XSIZE, YSIZE, and XST of the image.
The server process 1 sends a PRESCAN packet P1 including ART, YSTART, XSTEP, YSTEP, etc.
06 (arrow 131). XSIZE, YSIZE and the like are parameters that specify the size of an image to be read.

Since the server process 106 performs image coding and thinning during prescan, the PRESC
Upon receiving the AN packet P1, the encoding method, the bit rate at the time of encoding, and the thinning rate are set. Other information is the serial interface 2 as a prescan command.
From 15 to the copying machine 103 (arrow 132).

The copying machine 103 sets parameters in the scanner side image processing circuit 302 according to the information of the prescan command. When these parameters are set correctly, an OK response is sent back to the server process 106 (arrow 133). The server process 106 is transferred from the copying machine 103 to O.
When the K response is received, the OK packet P4 is sent back to the client process 105 (arrow 134). If the parameters are not set correctly, status information is sent back to the server process 106 instead of OK. When the copier 103 receives the prescan command, the image is set to 1
Read line by line. During the prescan, the video interface control circuit 216 of the scanner printer server 102 outputs the output timing signal group HSYNC.
(Horizontal sync signal) 110, VSYNC (vertical sync signal)
109, CLOCK (clock signal) 108, DATA
(Image data signal) 111 is set to high impedance,
The server 102 receives the input timing signal group HSYNC (horizontal synchronization signal) 110, VSY generated by the copying machine 103.
DATA (image data signal) 1 in synchronization with NC (vertical synchronization signal) 109 and CLOCK (clock signal) 108
Data is read from 11 (arrow 135) and written in the buffer memory 209.

At this time, the images are thinned out according to the thinning rate of XSTEP and YSTEP designated by the PRESCAN packet P1, and the image data is divided or combined into an appropriate size to obtain the DATA tag, the number of bytes of the packet, and the image. A plurality of DATA packets P consisting of data and
2 is configured and sent to the client process 105 of the host computer 101 (arrow 136).

The client process 105 of the host computer 101 takes out the thinned image data from the image packet P2 received from the server process 106 and displays it on a CRT (not shown). Copy machine 103
After sending all the image data normally, sends OK from the serial interface 307 to the server process 106 (arrow 137). The server process 106 is the copy machine 1.
If OK is received from 03, an OK packet P4 is sent to the client process 105 (arrow 138). After transmitting the OK packet P4 to the client process 105, the server process 106 waits for the next command packet from the client process 105. The client process 105 receives all the thinned image data,
Display on CRT and ask user which area of the image is actually read.

In response, the user specifies which area is actually scanned with a pointing device such as a mouse. It also gives an instruction as to whether the area designated by the user or the color character should be recognized. And
Instruct the client process 105 to start scanning.

Then, the client process 105
Image XSIZE, YSIZE, XSTART, YST
The SCAN packet P3 including ART or the like is sent to the server process 106 (arrow 139).

When the SCAN packet P3 is received, the server process 106 sends back an OK to the client process 105 in the same procedure as in PRESCAN when the information is set correctly (arrow 1310).

The server process 106 returns to the copying machine 10 again.
The image is read from 3 according to the parameters specified by the SCAN packet (arrow 1311).

In the case of reading a normal image, this image data is divided into appropriate sizes or combined, and DA
A plurality of DATA packets P5 including a TA tag, the number of bytes of the packet, and image data are constructed and sent to the client process 105 of the host computer 101 (arrow 1
312).

When the character recognition is designated, the character image data of the area or the color area is character coded by the character recognition means. At the same time, if there is an instruction for position detection or page description verbalization conversion, that information is added. And D
A plurality of DATA packets including an ATA tag, the number of bytes of the packet, and image data are constructed and sent to the client process 105 of the host computer 101.

The client process 105 of the host computer 101 takes out image data from the image packet received from the server process 106 and writes it in the disk one after another.

When the server process 106 normally sends all the images, it sends an OK packet P4 to the client process 105 (arrow 1313). Server process 10
6 sends an OK packet to the client process 106 and waits for the next command packet from the client process 105. When the client process 105 receives the OK packet from the server process 106, it waits for the next command from the user.

Here, when the destination of the character code information is a transfer to another party on the public line, there are two transfer routes according to an instruction from the host computer 101. If the server 102 has a communication control device, the server 102
From the host computer 10 directly to the public line
If the communication control device 1 is provided, the data is transferred from the server 102 to the host 101, and the client process 105 receives the data and transfers it on the public line.

[Printer Side] FIG. 14 is an explanatory diagram of a printing operation. The data exchange between the client process 105 of the host computer 101, the scanner printer server 102 and the copying machine 103 at the time of printing will be described below with reference to the drawing. To do.

First, the user is the host computer 10
1, the size of the image to be printed, the position of the image to be printed, the file name on the host computer 101 holding the image to be printed, etc. are designated. Also, parameters such as gamma conversion and masking conversion during printing are usually set to default values,
These can also be changed.

In this case, in FIG. 14, the client process 105 first sends a GAMMA packet P6 to the server process 106 in order to set the gamma table for gamma conversion at the time of image printing (arrow 14).
1). If the gamma table has already been set, there is no need to send it. Upon receiving the GAMMA packet P6, the server process 106 sends a gamma table setting command to the printer side of the copying machine 103 when the second byte of the GAMMA packet indicates a printer, and a scanner side when the second byte of the GAMMA packet indicates a scanner. Send via / F (arrow 14
2). In this case, since the printer of the copying machine 103 is used, the printer side gamma conversion table is set.

The copying machine 103 sets the LUT of the printer gamma conversion unit 502 according to the parameter of the gamma setting command. If it can be set normally, OK is sent back to the server process 106 (arrow 143). Server process 10
When receiving the OK from the copying machine 103, the No. 6 sends back the OK packet P4 to the client process 105 (arrow 144).

Next, the client process 105 sets M to set the masking table for image printing.
The ASKING packet P7 is sent to the server process 106 (arrow 145). If the masking table has already been set, it need not be sent. Upon receiving the MASKING packet P7, the server process 106 sends a masking setting command to the copying machine 103 (arrow 146). The copying machine 103 sets the parameters in the masking conversion unit 503 in the printer side image processing circuit 303 according to the parameters of the masking setting command. If it can be set normally, OK is sent back to the server process 106 (arrow 14
7). Upon receiving the OK from the scanner, the server process 106 sends the OK packet P4 to the client process 1
It returns to 05 (arrow 148).

When the parameters for gamma conversion and masking conversion have been set as described above, the client process 105 determines the XSIZE, YSIZE, XST of the image.
PRIN consisting of ART, YSTART, PAGE, etc.
Send T packet P8 to server process 106 (arrow 1
49).

Upon receiving the PRINT packet P8, the server process 106 receives the serial interface 21.
5 sends a print command to the copying machine 103 (arrow 14
10). When the copying machine 103 correctly receives the information of the print command and sets the parameter, OK is sent back to the server process 106 (arrow 1411). Server process 1
When 06 receives an OK from the scanner, the OK packet P
4 to the client process 105 (arrow 14
12).

When the client process 105 receives the OK packet, it reads the image from the specified file. The client process 105 divides the read image data into an appropriate size or combines them, and
A plurality of DATA packets P9 composed of an ATA tag, the number of bytes of the packet, and image data are constructed and sent to the server process 106 (arrow 1413). Server process 106
Then, the image data is taken out from the image packet P9 which is sent from the client process 105 of the host computer 101 and received, and the copying machine 1 is sent via the video I / F.
03 sequentially sends print data (arrow 141
4).

At the time of printing, the video I / F control circuit 216 of the scanner printer server 102 uses the sync signal group HSYNC (horizontal sync signal) 110, VSYNC.
(Vertical sync signal) 109, CLOCK (clock signal)
108 and the image data synchronized with this are output to DATA (image data signal) 111, and the copying machine 103 reads the data to be printed from the DATA (image data signal) 111 signal in synchronization with the synchronization signal group and prints it. .

When the copying machine 103 normally prints all the images, OK is sent to the serial interface 307.
To the server 102 (arrow 1415). When the server process 106 receives the OK from the copying machine 103, the O
The K packet is sent to the client process 105 (arrow 1416). After sending the OK packet to the client process 105, the server process 106 waits for the next command packet from the client. When the client process 105 receives the OK packet from the server process 106, it waits for the next command from the user.

In the above printing operation, in addition to the case where the image data is sent from the host computer 101 by the instruction of the client process 105, the scanner / printer server 102 reads the image data from the scanner of the copying machine 103 and converts it into a page description language. There are also cases in which the type and size of characters are changed and printed. Also, in the same way, receive character codes from the public line,
In some cases, the bits are expanded in the scanner printer server 102 and printed.

[Error] FIG. 15 is an explanatory diagram when an error occurs at the time of printing. The data exchange between the host computer 101, the scanner printer server 102, and the printer side 103 at the time of printing will be described below with reference to the drawing. .

Similar to the above printing, the client process 105 sets the gamma table and masking table for image printing. Next, the client process 105 uses the XSIZE, YSIZE, and XSTA of the image.
PRINT consisting of RT, YSTART, PAGE, etc.
Send the packet to the server process 106. Upon receiving the PRINT packet, the server process 106 sends a print command from the serial interface 215 to the copying machine 1.
03 (arrows 151 to 1510).

If the value of the parameter of the print command is incorrect, or if an error such as a failure to set it occurs or an error such as paper shortage occurs, the copying machine 103 outputs status information indicating the error status to the server process 106. It sends it back to (arrow 1511).

Upon receiving the status information from the copying machine 103, the server process 106 sets the status information to ST.
The ATUS packet P10 is converted and sent back to the client process 105 (arrow 1512).

Client process 105 is STATUS
When the S packet P10 is received, an appropriate message is output to the user according to the status to notify that an error has occurred. When an error such as a paper jam occurs in the printer, the C of the copying machine 103
The PU 306 immediately interrupts the print operation and transmits error status information to the server process 106 via the serial interface 307.

In the server process 106, the printer side 1
When the status information from No. 03 is received, this status information is sent to the client process 105 as a STATUS packet P10, and the next command is awaited.

Client process 105 is STATUS
When the S packet P10 is received, an appropriate message is output to the user according to the status to notify that an error has occurred.

[Character Recognition Server] In some cases, only the character recognition means and the automatic translation means in the scanner printer server 102 are used from the host 101. In this case, the host computer 101 side cuts out character image data in a desired area in advance, and the client process 1
05 sends the character image data to the server process 106.
In the server 102, the same processing as when the image is read by the copying machine 103 is performed, and the character code is returned to the host computer 101 via the network 104. In this case, the copying machine 103 is not involved in the processing, and the host 101
The message sent from the server to the server 102 is processed by the server 102, and no instruction is given to the copying machine 103 to perform the processing.

The following effects can be obtained by using the copying machine configured and functioning as described above and its server system.

(1) The copier can be combined with a scanner printer server capable of character recognition to form a shared resource on the network. That is, the network user
In addition to the copier itself, the server functions can be shared.

(2) Since a character code is obtained by reading a character image document with a scanner, the amount of data is reduced, the load on the network is reduced, and the processing speed in the computer is increased and the operability is improved.

(3) A character image original read by a scanner can be automatically converted into a page description language.

(4) It is possible to read a character image original containing a color and take out only a character of a designated color.

(5) The type and size of characters of the read character image original can be changed and printed.

(6) The read character image original can be recognized as a character, converted into a character code, and further automatically translated.

(7) The host computer on the network can use only the character recognition function of the server.

(8) Character image An original document can be recognized as a character code and sent as a character code to a public line, or the character code can be received and printed.

(9) PDL the image read by the scanner
By converting to, the quality of the image data can be reduced, the memory required for the image data can be reduced, and the time required to transfer the image data can be shortened.

In this embodiment, the case where the bus type Ethernet 104 is used for the network has been described, but the network may be of any type. For example, FDDI using a token ring or an optical fiber is also possible as a network means.

Further, although the case where the serial communication is used to communicate the data such as the command, the parameter and the error between the scanner printer server 102 and the copying machine 103 has been described in the present embodiment, it is not limited to the serial communication. Any two-way communication interface can be used.

In this embodiment, serial communication is used to communicate data such as commands, parameters and errors between the scanner printer server 102 and the copying machine 103, and a video interface is used to communicate image data. It is also possible to communicate information such as commands, parameters and errors and image data on the same communication path by using an interface capable of bidirectional communication such as SCSI and GPIB without being restricted by these communication interfaces. .

Also, in this embodiment, the image is dot-sequential YC.
In the case of the rCb format, the image is sent by being coded by the ADCT coding method, but not limited to the ADCT coding method, any coding method can be used. As a result, encoding is also performed for images other than the dot-sequential YCrCb format,
Images can be compressed and transmitted.

Further, although the buffer memory 209 for several lines is used in the server 102, the memory for one screen can be provided to further increase the speed.

Further, in the present embodiment, the scanner unit 310 of the copying machine 103 thins out the image during the pre-scan and sends the image during the main scan without thinning it out.
When pre-scanning a color image, (1) send only single color components (2) send with thinning out (3) send with encoding (4) send with reduced number of image gradations (5) drop image resolution It is also possible to combine sending by sending.

Further, in this embodiment, when the image is read, scan converted and transferred, the sensor 301 scans three times. However, it is also possible to read the image only once, store the image data in the hard disk, read the image from the hard disk three times, and replace it with scanning. As a result, only one mechanical scan is required, and the speed can be increased.

It is also possible to scan and convert the first color at the same time as storing in the hard disk and read from the hard disk for the remaining two times.

In this embodiment, the binarizing unit 505 binarizes the image for the binary printer engine. This can be dealt with in the same manner, and in the case of an N-value printer engine, it can be easily dealt with by providing an N-value conversion unit.

The present invention may be applied to a system composed of a plurality of devices or an apparatus composed of one device. Further, it goes without saying that the present invention can be applied to the case where it is achieved by supplying a program to a system or an apparatus.

[0346]

[Other Embodiments] FIG. 20 is a diagram of another embodiment. The functions provided are the same as those in the previous embodiment, but the copying machine 181 in the figure is an apparatus in which the scanner printer server 102 is incorporated in the copying machine 103 in the first embodiment, and the copying machine 181 performs the functions of the server 102. Also fulfills.

Further, although the copying machine has been described in this embodiment,
The present invention is not limited to a copying machine, but can be similarly applied to any device provided with a scanner such as a facsimile device.

[0348]

As described above, the image processing apparatus and method according to the present invention can effectively use a scanner / printer as a shared resource of a network, and can also share various image processing services.

[Brief description of drawings]

FIG. 1 is a configuration diagram of a scanner printer system embodying the present invention.

FIG. 2 is an internal configuration diagram of a scanner printer server.

FIG. 3 is a configuration diagram of a scanner printer.

FIG. 4 is a block diagram of an internal image processing circuit on the scanner side.

FIG. 5 is a block diagram of the internal image processing circuit of the printer.

FIG. 6 is a configuration diagram of a character recognition circuit.

FIG. 7 is an explanatory diagram of character segmentation and feature extraction.

FIG. 8 is a block diagram of an ADCT interpreter coding.

FIG. 9 is a block configuration diagram of a PDL conversion circuit.

FIG. 10 is a block diagram of a translation circuit.

FIG. 11 is a diagram showing VSYNC, HSYNC, and raster scan.

FIG. 12 is a diagram showing Video I / F timing.

FIG. 13 is a scanner state transition diagram.

FIG. 14 is a printer state transition diagram.

FIG. 15 is an error state transition diagram.

FIG. 16 is a diagram of a server command packet.

FIG. 17 is a diagram of a server command packet.

FIG. 18 is a diagram of a scanner command packet.

FIG. 19 is a diagram of a scanner command packet.

FIG. 20 is a configuration diagram of another embodiment.

FIG. 21 is a diagram of an example of an image in which characters and images are present.

FIG. 22 is an explanatory diagram of image thinning processing.

FIG. 23 is an explanatory diagram of a process of dividing an image into rectangles.

FIG. 24 is a diagram of an example of recognizing a region.

FIG. 25 is an explanatory diagram of a process of recognizing a direction in which characters are assembled.

FIG. 26 is a diagram of an example in which character data is converted into PDL.

FIG. 27 is a diagram of an example in which image data is converted into PDL.

FIG. 28 is a diagram showing an example of converting graphic data into PDL.

FIG. 29 is a flowchart of a process of converting an image into PDL.

FIG. 30 is a flowchart of a process of converting an image into PDL.

FIG. 31 is a flowchart of a process of converting an image into PDL.

FIG. 32 is a flowchart of a process of converting an image into PDL.

[Explanation of symbols]

101 ... Host computer, 102 ... Scanner printer server, 103 ... Scanner printer, 104 ... Ethernet, 105 ... Client process, 106 ... Server process, 107 ... SPCOM (scanner printer serial command signal), 108 ... CLOCK (clock signal, 109 ... VSYNC (vertical synchronization signal), 110 ...
HSYNC (horizontal synchronizing signal), 111 ... DATA (image data signal), 201 ... CPU, 202 ... PROM, 2
03 ... RAM, 204 ... Serial interface, 2
05 ... Parallel interface, 206 ... Ethernet controller, 207 ... Communication control unit (CCU), 2
08 ... Dual port RAM, 209 ... Buffer memory for image storage (for several lines), 210 ... Character recognition circuit, 21
1 ... Image encoding / decoding circuit (compression / decompression), 212 ...
PDL conversion circuit, 213 ... Automatic translation circuit, 214 ... Sub CPU (for scanner / printer interface), 21
5 ... Serial interface (for SPCOM), 21
6 ... Timing control circuit for Video I / F, 217
... Serial interface port, 218 ... Parallel interface port, 219 ... Ethernet port, 220 ... Public line I / F port, 221 ... Vide
o I / F command port, 222 ... Video I /
F data port, 223 ... Main CPU bus (local bus or VME bus, etc.), 224 ... Sub CPU bus, 301 ... Scanner image reading sensor, 302 ... Scanner side image processing circuit, 303 ... Printer side image processing circuit, 304 ... Print head, 305 ... Scanner printer control CPU, main memory, 306 ... Scanner printer control CPU, 307 ... Serial interface, 308 ... Video I / F timing control circuit, 309 ... Scanner printer internal image data bus, 4
01 ... Shading correction, 402 ... Resolution converter, 4
03 ... Scanner color conversion, 404 ... Scanner gamma conversion unit, 405 ... Level conversion unit, 501 ... Printer color conversion unit, 502 ... Printer gamma conversion unit, 503 ... Masking conversion unit, 504 ... Black generation, undercolor removal unit, 505 ... Binarization unit, 601 ... DSP (Digital Signal) for character recognition control
Processor), 602 ... Dual port RAM, 603 ...
Feature extraction circuit, 604 ... Identification circuit (dictionary), 605 ... Local bus (for character recognition), 801 ... R (Red) component input section, 802 ... G (Green) component input section, 804
... B (Blue) component input unit, 804 ... Color conversion unit, 80
5 ... Y (luminance) component input unit after color conversion, 806 ...
Cr (chrominance) component input unit after color conversion, 807 ...
Cr (chrominance) component input unit after color conversion, 808 ...
8 * 8 DCT (Discrete Cosine Transform) processing unit, 8
09 ... Linear quantizer, 810 ... One-dimensional predictor, 811 ...
Huffman coding unit, 812 ... Zigzag scanning, image data input unit, 813 ... Two-dimensional Huffman coding unit, 814 ... Multiplexing unit, 901 ... PDL interpreter CPU, 90
2 ... Main memory for CPU, 903 ... ROM for storing program, 904 ... Local bus, 1001 ... CPU for automatic translation control, 1002 ... Main memory for CPU, 10
03 ... ROM for program storage, 1004 ... Clause processing circuit, 1005 ... Identification circuit (translation dictionary), 1006 ... Document character code input / output dual port RAM, 1007
… Local bus

Claims (10)

[Claims] 1. An input unit for inputting image data, a sorting unit for sorting the image data input by the input unit according to the type of image included in the input unit, and a sorting unit according to the type of image recognized by the sorting unit. An image processing apparatus, comprising: an encoding unit that encodes image data according to the present invention; and a conversion unit that converts the image data and the code data obtained by the encoding unit into a page description language. 2. The image processing apparatus according to claim 1, wherein the partitioning unit separates the image data into an image area, a graphic area, and a character area and adds position information of each. 3. The encoding means has a character recognizing means, and the image data of the character area divided by the dividing means is recognized by the character recognizing means and converted into a character code. The image processing device according to item 1. 4. The encoding means has a figure recognizing means, and the image data of the figure area divided by the dividing means is recognized by the figure recognizing means to form a figure code. The image processing device according to item 1. 5. The image processing apparatus according to claim 1, further comprising means for compressing image data. 6. The image processing apparatus according to claim 2, wherein the position information includes a character size, a pitch, and a position. 7. The image processing apparatus according to claim 2, wherein the position information includes a line type of a figure, a line thickness, a line length, and a position. 8. The image processing apparatus according to claim 2, wherein the position information includes a size, a position, and an outer shape of the image. 9. The image processing apparatus according to claim 1, further comprising communication means for performing communication via a communication network, wherein the output of the conversion means is transmitted to the communication network by the communication means. 10. An input step of inputting image data, a classification step of classifying the image data input by the input step according to the type of image included therein, and a classification step according to the type of image recognized by the classification step. An image processing method, comprising: an encoding step of encoding image data according to the above method; and a conversion step of converting the image data and the code data obtained by the encoding step into a page description language.